Public housing development (PHD) residents typically lack the resources and information necessary to maintain good oral health. Children living in PHDs are at increased risk for oral health disparities as the social conditions are thought to hinder the flow of useful and actionable health information, and undermine exemplary health behavior practices. Current oral health research tools are insufficient to accurately characterize PHD resident interactions and to understand how the social relationships of caregivers may be effectively leveraged to improve their children's oral health. To date, disparities research has no adequately accounted for the heterogeneity of factors interacting in a web of influences nor permitted practical measurement of them. Network Science is ideally suited for investigating complex systems. Social networks have been conceived as key linking mechanisms characterizing interactions within a system. We will use Network Science methods to design and implement a network intervention, guided by the diffusion of innovations theory. The intervention will focus on both interactions between residents of PHDs, and the upstream influences operating within and among networks that affect more proximal pathways to health status. Our proposed multi-level intervention is aimed at producing individual, network and community-level outcomes. Our central hypothesis is that controlled dissemination of information targeting changes in health behaviors through social networks will accelerate the uptake of knowledge, support the adoption of positive behavior change, and improve oral health status of caregivers and their young children. During the formative UH2 phase, a multidisciplinary team of investigators will identify and characterize the structural properties of PHD members' networks that shape the flow of health information, resources and support. We will rely on community-based participatory research to maximally leverage our established, long-term links with community partners. In the UH3 phase, we will conduct an intervention to determine whether and how networks may be used to promote and accelerate health behavior changes. We will target caregivers of children ages 1-5 years, with 18-24 months' follow-up to ascertain the effectiveness of network activation through disseminated messages to acquiring knowledge and promoting behavior change with respect to healthy dietary choices, particularly reductions in sugar-sweetened beverage consumption. The research plan represents a major methodological improvement to the current state of the science as it addresses the complex systems' nature of oral health disparities. It will enhance our understanding of oral health disparities through bette characterization of the behaviors, social interactions (including interactions with health systems) and community features related to the information PHD residents utilize to make sense of, and act on, their health. Similar Network Science approaches in other fields have shown increased sustainability of behavior change over and above that of individualized efforts.